Ask why housing is expensive and you will usually get an answer that names a culprit. Landlords. Developers. Investors. Zoning boards. Interest rates. Immigration. Airbnb. Each answer has a constituency and each has a grain of truth, which is exactly why the debate never resolves. Housing Unpacked starts from a different place. Before asking who is responsible for the price of housing, it is worth understanding what the price is made of.
This article is the foundation for everything else we publish. It explains the single idea that organizes the publication: the cost of a home is the sum of every input required to create and operate it, divided by the number of homes that share those inputs, and translated into a monthly figure by the cost of money. Every requirement, fee, delay, rate change and design choice enters that sum. Most of them are individually reasonable. Together they determine what a new home has to rent for, and whether it gets built at all.
The question most housing debates skip
Housing is unusual among consumer goods in that most people never see how it is produced. A renter sees a monthly number. A buyer sees a listing price. Neither sees the pro forma, the spreadsheet in which a developer or lender lays out every cost of a project and every dollar it is expected to earn. The pro forma is where the price of housing is actually decided, long before a lease is signed, and it is where most of the controversies in housing policy quietly play out.
The pro forma answers one question: does this building work? “Work” has a specific meaning. It means the building, once complete and leased, will produce enough revenue to pay its operating costs, service its debt, and return enough to the people who put up the equity that they would rather have done this than something else with their money. If the answer is yes, the building gets built. If the answer is no, it does not, and the land stays a parking lot, a warehouse, or a smaller building than it could have been.
That binary is the most important thing to understand about housing costs. A regulation does not make an apartment cost $50 more a month. It makes the apartment either feasible at $50 more a month, or, if the market will not pay that, not feasible at all. The cost of an unbuilt building is invisible, which is why it is so easy to ignore.
Housing is a stack of inputs
Every new home requires the following, in roughly the order a developer encounters them. None of them is optional, and each one has a price.
- Land, purchased at a price that reflects what else could be built on it under current rules.
- Entitlement: the permissions to build, which take time and professional fees to secure, and which may come with conditions.
- Hard costs: labor and materials for the building, the site, and the parking.
- Soft costs: architecture, engineering, legal, surveys, permits, insurance during construction, marketing, and the developer’s overhead.
- Government requirements: impact fees, utility connection charges, permit fees, and sometimes mandated affordable units, public improvements or design standards.
- Financing: a construction loan whose interest accrues while nothing is being earned, followed by a permanent loan that must be serviced every month for decades.
- Equity: the portion of the cost that lenders will not fund and that investors provide in exchange for a return that compensates them for the risk.
- Operations: management, maintenance, payroll, repairs, marketing and turnover, every year the building exists.
- Property taxes, insurance and utilities: recurring costs that the owner does not control and that have risen faster than rent in many markets.
- Reserves: money set aside because roofs, boilers and parking lots wear out.
- Vacancy: an allowance for units that sit empty between tenants and rent that goes uncollected.
Some of these are one-time capital costs. Others recur every year. The pro forma converts them all into the same unit, dollars per month per home, which is the number a renter eventually sees.
The arithmetic that decides whether a building gets built
The conversion from capital cost to monthly rent is the piece most people have never seen written down, so here it is. We use the same method throughout Housing Unpacked and in our Rent Impact tool.
Annual debt service = (Capital cost × Loan-to-cost) × Mortgage constant
Annual equity return = (Capital cost × (1 − Loan-to-cost)) × Required equity yield
Required NOI = Annual debt service + Annual equity return
Required revenue = (Required NOI + Operating costs) ÷ (1 − Vacancy)
Required rent = Required revenue ÷ Units ÷ 12NOI is net operating income: revenue after operating costs but before debt and returns. The mortgage constant is the annual payment on a loan as a share of the amount borrowed.
Under the financing assumptions we use by default, which are stated in full below, each dollar of capital must earn about 7.73 cents a year: 65 cents of it is borrowed and costs 7.585 cents a year in principal and interest, and 35 cents of it is equity that requires an 8 percent cash yield. Divide by twelve, gross up for a 5 percent vacancy allowance, and the result is a rule of thumb worth remembering.
$1,000 × 0.0773 ÷ (1 − 0.05) ÷ 12 = $6.78. Illustrative; the exact figure depends on the financing assumptions.
Why every dollar of cost shows up as rent
A capital cost is paid once but must be recovered continuously, because the money used to pay it was either borrowed and must be repaid with interest, or invested and must be compensated. There is no third option in which a cost is simply absorbed. A developer who agrees to a cost that rent will not recover is a developer whose lender declines the loan or whose investors decline the deal. This is not a moral observation about developers. It is a description of how any capital project, public or private, is paid for.
The rule of thumb scales. A requirement that adds $10,000 per home adds about $68 a month. A requirement that adds $3 million to a 120-home building, or $25,000 per home, adds about $170 a month. A twelve-month delay that adds carrying costs and pushes revenue a year further away adds its own increment, which we work through in What a Year of Development Delay Costs.
Costs do not set rent. The market does. So why do costs matter?
Here is the nuance that trips up both sides of the affordability debate. Developers cannot charge more than the market will pay, and they cannot charge less than their costs require. Rent for a new building is set by what comparable homes in the area rent for. Cost determines only whether the project can be built at that rent.
It follows that a new cost does not immediately raise the rent on existing buildings. A property tax increase, a new fee, or a higher interest rate lands first on the pro forma of the next building, not on the lease of the current one. If the pro forma no longer works, the next building is not built. The effect on rents arrives later and indirectly, through supply: fewer new homes are added, the existing stock has less competition, and rents across the market drift upward relative to where they would otherwise have been. The mechanism is slow, diffuse and therefore easy to deny, but it is the main way costs reach renters.
The feasibility gap
We call the distance between the rent a project requires and the rent the market will pay the feasibility gap. When required rent is below market rent, projects are proposed, financed and built, and the added supply pushes back on market rent. When required rent is above market rent, nothing new is built for that segment of the market, and the gap can only close from one of two directions: costs fall, or rents rise. In most high-cost regions over the past decade the gap has closed from the second direction.
This is why Housing Unpacked spends so much time on new construction even though most people live in older buildings. The price of new housing is the price at which supply responds. Whatever sets that price, sets the ceiling on affordability for everyone.
The inputs, one at a time
Land
Land is priced by what can be built on it. A parcel that permits sixty homes is worth more than one that permits six, and a parcel that permits nothing new is worth what it earns today. This means zoning and density rules do not just determine how many homes can exist on a site; they determine the land cost each of those homes must carry. Cutting permitted density by a quarter does not cut the land price by a quarter in the short run, because the seller expects what the market has been paying. It raises the land cost per home.
Construction
Hard costs are typically the largest single input, and they are the one input that is close to the same everywhere for a given building type, because labor and materials trade in regional and national markets. Building codes, structural requirements, energy standards, accessibility requirements and parking all live here. So does the choice between wood-frame garden buildings, mid-rise with a concrete podium, and high-rise steel and concrete, each with a cost per square foot that is a multiple of the last. We explain why the cheaper options are so often unavailable in Why Can’t Developers Build Cheaper Apartments?.
Soft costs and time
Design, engineering, legal work, permits, studies and insurance are individually small and collectively significant. Many of them scale with time rather than with the building. A project that takes three years to approve carries three years of professional fees, option payments and staff time before a shovel touches the ground, and all of it is capitalized into the cost the rent must recover.
Financing
The cost of money enters twice. During construction, interest accrues on a loan that is funding a building with no tenants, and that interest becomes part of the project cost. After completion, the permanent loan is serviced every month for decades. Because the loan is a fixed obligation and rent is not, lenders require a cushion, measured by the debt service coverage ratio, which further raises the rent a project must be able to show on paper. We quantify the sensitivity in What Happens When Interest Rates Rise 1%?.
Government requirements
Impact fees, utility connection charges, permit fees, dedications, required public improvements, inclusionary units, parking minimums and design mandates each have a purpose and a price. Some purchase real public goods. Some substitute for revenue the public would otherwise have to raise from taxpayers who vote. All of them are paid, in the first instance, by the people who will live in the building, because there is nowhere else in the pro forma for them to go. The honest question is never whether a requirement is worth having. It is whether it is worth its cost in rent, and whether the people paying that rent are the right people to bear it.
Operating costs, taxes and insurance
A building must be run. Management, maintenance, staff, repairs, marketing and turnover recur every year, and so do property taxes, insurance and the utilities the owner pays. These are cash costs, not capital costs, so they pass into required rent one-for-one, grossed up only for vacancy. When property insurance premiums double in a coastal market, the effect on required rent is immediate and exact.
Returns
Two kinds of return appear in every project. Investors who supply equity require a yield, because their money is at risk and could be deployed elsewhere. Developers who assemble the project earn a developer fee and typically a share of profits if the project outperforms. How large these returns are, how they compare with the risk, and what would happen to rent if they were smaller are fair questions, and we answer them with numbers in Does Developer Profit Make Housing Unaffordable?. The short version is that returns are a real component of rent and a smaller one than most people assume, because most of the capital in a building is borrowed and the lender’s return is fixed.
Why “who is to blame” is the wrong question
Once the inputs are laid out, the blame framing starts to look unhelpful. A developer who charges the maximum the market allows is doing what a lender requires. A city that imposes a fee is paying for infrastructure it cannot otherwise fund. A neighbor who opposes density is protecting the value of the largest asset they own. An investor who demands an 8 percent yield is comparing it with what a bond pays. Each actor is responding rationally to incentives, and the outcome nobody chose is a required rent that fewer and fewer households can pay.
Housing affordability is not a price someone sets. It is the residual of a hundred decisions made by people who never see the total.
The useful question is not who to blame but which inputs are large, which are movable, and what each one buys. That is a question that can be answered with a spreadsheet, and answering it is what this publication is for.
The tradeoffs are real on both sides
It would be easy to read the argument so far as a case for stripping out every requirement. That is not the argument. Parking has value to the people who use it. Energy codes reduce utility bills and emissions. Impact fees build the roads and pipes new residents will use. Inclusionary units house people the market would not. Design review can produce better streets. Each of these is a legitimate public purpose, and a publication that pretended otherwise would not deserve to be trusted.
What is not legitimate is pretending the purposes are free. Every one of them is paid for in required rent by the future residents of the building, who are, almost by definition, not in the room when the requirement is adopted. The point of quantifying is to let a community decide with its eyes open: this rule costs about this much per month, per home, and here is what it buys. Some rules will survive that test easily. Others will not. Housing Unpacked does not take a position on which is which. We take a position on doing the math.
How to read Housing Unpacked
Every quantitative article on this site follows the same discipline. Assumptions are stated in a labeled block. Every number is either an assumption or is derived from one, with the derivation shown. Illustrative projects are labeled as illustrative. Sources are listed, and when we have not yet verified a source we say so rather than cite it. A “What would change this” block lists the conditions under which our conclusion would be different. Where a figure depends on financing assumptions, you can change them yourself in the Rent Impact tool.
If you are new to development, start with Housing 101, an eight-lesson sequence that builds from how housing gets built to how supply and demand set rent. Then read Why a New Apartment Costs $2,000 a Month, which applies the arithmetic above to Reference Project A, our garden building, and shows where every dollar of required rent comes from. Articles about capital-intensive buildings use Reference Project B, a mid-rise over a structured garage, which costs more and therefore requires more; each article names its project in its assumptions, and both are defined on our Methodology page. From there, each article in the Why We Can’t Build series isolates one input and asks the only question that matters: what does this add to rent?
This article models no particular building. It uses only the shared financing assumptions below, which both of the site's reference projects share: Reference Project A (the garden project), 120 wood-frame units with surface parking, $24,265,760 of cost and a required rent of $1,996; and Reference Project B (the podium project), 120 units in a mid-rise over a structured garage, $30,000,000 of cost and a required rent of $2,415. Why no project here: the argument is about the shape of the arithmetic rather than any one building, and the two projects show that the same arithmetic produces different rents when the building changes. Both are defined side by side on our Methodology page.
We assume 65 percent of capital cost is funded by a permanent loan (loan-to-cost of 0.65) and 35 percent by equity.
We assume the permanent loan carries a 6.5 percent fixed interest rate with 30-year amortization, which produces a mortgage constant of 7.585 percent (annual payment as a share of the loan).
We assume equity investors require an 8 percent annual cash-on-cash yield, before any return of their capital.
We assume a 5 percent vacancy and credit-loss allowance.
Blended annual capital charge: 0.65 × 0.07585 + 0.35 × 0.08 = 0.0773, or 7.73 cents per dollar of capital per year.
Rent added per $1,000 of capital cost per unit: $1,000 × 0.0773 ÷ 0.95 ÷ 12 = $6.78 per month. Per $3 million on 120 units ($25,000 per unit): $169.53 per month.
These are the default assumptions of our Rent Impact tool. They are illustrative, not a forecast of any market, and every article that uses them says so.
This article is a framework, not an empirical study. The one calculation it contains, the rent added per $1,000 of capital cost, follows the shared rent-math method described on our Methodology page and used by every calculator on the site. The method converts a capital cost into an annual obligation (debt service on the borrowed share plus a required yield on the equity share), grosses that obligation up for vacancy, and divides by units and months.
The method is deliberately simple. It ignores income taxes, depreciation, loan fees, appreciation, and the fact that equity investors are usually repaid at sale rather than from rent. Each of those omissions is discussed in the articles that depend on it. The simplicity is a feature: a reader can reproduce every figure with a calculator.
It follows that this piece claims no external data. Loan terms, required returns and construction costs move with the market, and we have not gone out and measured any of them; the figures here exist to show the shape of the relationship, not its level in any particular city in any particular year. A reader who wants the answer for a real place should run the same arithmetic on local numbers, which is the whole point of showing the arithmetic.
If lenders funded a larger share of cost at lower rates, or investors accepted lower yields, the capital charge per dollar would fall and every cost input would translate into less rent. Public or subsidized financing is one version of this, examined in Could Government Finance Housing More Cheaply?.
If a market’s rents were far above required rent, added costs would reduce developer and land-owner profit before they reduced supply. The framework applies most directly where the feasibility gap is narrow, which describes most markets where new construction is marginal.
If new supply had no measurable effect on market rents, the indirect channel described here would not operate. The weight of published evidence points the other way, but the size of the effect varies by market and is a fair subject for scrutiny.
If a cost input generated offsetting revenue (for example, parking that tenants pay for separately), its net effect on required rent would be smaller than its gross effect. Where this is plausible we model it explicitly.
What this means
The price of housing is not a mystery and it is not a conspiracy. It is a sum. Land, construction, time, money, rules, operations and returns, added together, divided by the homes that share them, and translated into a monthly figure by the cost of capital. Each input is defensible on its own terms. Each one is also paid for by the people who will live in the building, and if the sum exceeds what they can pay, the building is not built and the shortage gets a little worse.
Understanding this does not tell you what housing policy to support. It tells you what any policy will cost and who will pay for it, which is the information a community needs in order to choose. Everything Housing Unpacked publishes is an attempt to make that sum visible, one input at a time.